Krabbe leukodystrophy (KD) is a fatal neurodegenerative lysosomal storage disorder caused by deficiency of galactosylceramidase (GALC) that affects both central and peripheral nervous systems. KD manifests in infants in the first few months of life and presents with severe irritability, muscle rigidity and motor deterioration, which quickly progresses to overall clinical decline and death within months. Unfortunately, there is no cure for KD. Our limited understanding of the pathogenesis is based on clinical data and on the spontaneous twitcher mouse model. Hematopoietic stem cell transplantation (HSCT) partially attenuates the course of KD only if performed before the onset of symptoms, presumably because stem cell derivatives secrete GALC that is uptaken by myelinating glia via the mannose-6-phosphate receptor, so called cross-correction. However, it is not clear how efficiently cross-correction happens in vivo, if only myelin-forming glia need to be corrected and at which developmental stage. Furthermore, accumulation of the lipid psychosine due to GALC deficiency contributes to KD by killing myelin-forming glia and neurons, but the relative importance of psychosine, its origin and the sequence of pathogenic events is unclear. We recently developed a conditional Galc floxed mouse and found that: 1) A KD-like phenotype similar to twitcher is obtained when Galc ablation is induced ubiquitously [Galc-iKO] at P4 or before. In contrast, induction at P6 or later significantly delayed the phenotype and prolonged survival (~25 days). Galc deletion before P4 caused severe developmental brainstem problems that were milder if deletion was induced after P6, indicating that GALC may be required for brainstem development; 2) Oligodendrocyte (OL)-specific Galc conditional knockout [Galc-CKO] results in a phenotype that includes tremor, wasting, kyphosis, motor defects, demyelination and mild axonal degeneration, but that is not as severe as Galc- iKO mice, suggesting that Galc deficiency in OLs may be not sufficient to trigger a complete KD phenotype; and 3) GALC uptake is less efficient in Galc-null cells in vitro, and surrounding WT cells provides minimal GALC to Galc-deficient OLs in vivo, indicating inefficient cross-correction of GALC. We propose 3 Aims to determine; 1) if GALC has an early role in brain development and contributes to mortality, 2) the key cells in the progression of KD pathology, and 3) the efficiency of cell-specific cross-correction of GALC. By combining a series of in vitro experiments with the comparison of cell-specific, time?specific and constitutive deletion of Galc in vivo, we will test the following 3 hypotheses that derived from our preliminary data and from the clinical experience: 1) GALC has specific developmental roles during critical periods including P4-6 in Krabbe mice; 2) any brain cell can produce psychosine or be the target of toxicity; 3) HSCT fails to cure KD due to inefficient cross-correction of GALC. Our results will help to understand the disease mechanisms of KD and the limitations of HSCT, which will allow the development of better therapies for KD and similar lysosomal, neurodegenerative and demyelinating diseases.